This PhD thesis identifies and analyses the Christian edifices of late antiquity in the province of Tarraconensis. The appearance of Christian buildings in this area during antiquity is neither a linear nor a uniform process, but instead develops according to a wide diversity of forms and circumstances. The study of this phenomenon is key to understanding early Christianity on the Iberian peninsula, as the transformations affecting both the countryside and the old Hispanic urban centres from the Peace of the Church onwards are varied and rich in both forms and nuances: the christianisation of settlements, the building of episcopal communities, newly built churches both inside and outside the city walls, the celebration of former places of private worship (domus ecclesiae), the christianisation of pagan temples with or without building alterations, the conversion of public facilities such as, for example, baths or civil basilicas, the construction of religious buildings on symbolic sites, like those within the arenas of amphitheatres, and so on and so forth. The truth is that the few surviving remains of Christian buildings of late antiquity in Tarraconensis are clearly only a very small part of the Christian edifices and churches that almost certainly existed before the Muslims arrived in the Iberian peninsula. Accordingly, this thesis has enlarged this catalogue by pursuing two main lines of action: 1) A thorough trawl through written sources contemporary to the first four centuries of Christianity, and 2) an analysis and dissection of a number of Roman archaeological sites and medieval churches, which despite the absence of direct information regarding the purpose of this study allow identifying new aspects of Christian edifices of late antiquity based on the observation of a series of significant common features (presence of necropolis, liturgical elements, Christian inscriptions, saints’ names, dedications, etc.). Based on a prior pool of fewer than twenty known edifices or sites, this study has managed to extend the catalogue to include as many as 252 examples, of which 87 have been rated as definite, 81 as very probable, 35 as probable and 49 as possible. Using the annotated and critical inventory that has been compiled, a series of significant common features can be observed and a number of conclusions are reached that would not have been possible to foresee beforehand.